MIT City Form Lab collaborative project
My role: Key collaborator in geospatial data engineering, spatial computation, model validation, and research visualization.
This is a lab project. It is presented separately from my independent dissertation research, Seeing HAI.
Project overview
Walking is central to urban travel, but cities rarely have consistent pedestrian counts for an entire street network. NYC Walks developed a citywide model of foot traffic during six peak travel periods in New York City. The model estimates pedestrian volumes across a network of 315,577 connected sidewalk, crosswalk, and footpath segments.
The work uses urban network analysis to connect trip origins and destinations through the pedestrian network, then calibrates the resulting estimates against observed counts. The citywide model provides a common baseline for comparing pedestrian demand, street classifications, infrastructure needs, and exposure to traffic risk.
My contribution
Within the broader City Form Lab team, I:
- contributed to large-scale geospatial data engineering and spatial computation;
- tested and validated workflows used to detect and assemble sidewalk and crosswalk networks;
- produced pedestrian-flow figures and visual comparisons between estimated volumes and observed counts; and
- contributed analyses and revisions to the published paper.
These contributions are narrower than the full project. The research design was conceived by Andres Sevtsuk, and the modeling, calibration, safety analysis, and writing were distributed across the author team.
Findings and planning applications
The published analysis identifies outer-borough streets whose estimated foot traffic is comparable with parts of central Manhattan but whose official classifications give them less pedestrian priority. It also compares crash counts with estimated pedestrian exposure and finds that some of the highest exposure-adjusted risks occur outside Manhattan.
These results show how a citywide pedestrian baseline can help agencies identify places where infrastructure classification, investment, and observed demand do not align.
City-scale pedestrian network analysis in New York City.
Publication
Sevtsuk, A., Basu, R., Liu, L., Alhassan, A., & Kollar, J. (2026). “Spatial Distribution of Foot Traffic in New York City and Applications for Urban Planning.” Nature Cities, 3, 136–145. Paper
The project was also presented with Sidewalk Ballet in the City Form Lab installation at the 2025 Venice Architecture Biennale.